Trends in the FDIS

[Marc]

From: ISO Standards Discussion
Date: Mon, 28 Aug 2000 15:49:52 -0500
Subject: Re: FDIS 9001:2000 "Trends" /Green/Paten

From: Mike Paten

> From: Joseph Green "To anyone lurking out there that
> knows FDIS content;
>
> Having observed the CD1, CD2, DIS revisions (with a great deal of joy)
> I'm curious if the word "Trends" remains as content in the FDIS
> version? What do others see as it's significance to "processes" and
> "products"? Where minimalist organizations have shown no "visible",
> "measurable" improvement; how will you "RAB" auditors respond to
> corporate smoke vs. valid corporate numbers?
>
> Love to hear comments prior to Oct. 2, when I'm charged with speaking
> on "my opinion" to the 9K2K revisions."

According to the DIS and/or FDIS - measurement requires (or infers) data collection of anything that represents desirable or undesirable product or process outcomes - while analysis requires (or infers) that these measurements must be viewed over time to assign cause and determine the need for corrective action and/or improvement. The word "trend" refers to any data viewed over time.

"Smoke" is "selective" measurements that show the "desired trend" - usually requiring no action (or action the company was going to take anyway). "Real trend" data represents the process "cycle" over a period that is likely to capture most outcomes (desired or otherwise) and will almost always result in some action - because continuous improvement never ends. Hence, the need to prioritize improvements on the basis of greatest need and/or opportunity.

Auditors (RAB or otherwise) will only be able to distinguish "smoke" from "real trend" data based on their experience with the processes they are auditing - which for the most part will not change with issuance of the forthcoming revisions to the ISO 9000 standards.

Although I would guess that even inexperienced auditors will be able to detect the "smell of smoke" - and ask probing questions accordingly.

Mike Paten

[Marc]

From: ISO Standards Discussion
Date: Tue, 29 Aug 2000 09:49:23 -0500
Subject: Re: FDIS 9001:2000 "Trends" /Green/Kozenko

From: David M. Kozenko

From: Joseph Green
> Having observed the CD1, CD2, DIS revisions (with a great deal of joy)
> I'm curious if the word "Trends" remains as content in the FDIS version?
>
> What do others see as it's significance to "processes" and "products"?
>
> Where minimalist organizations have shown no "visible", "measurable"
> improvement; how will you "RAB" auditors respond to corporate smoke vs.
> valid corporate numbers?

I see no difference between ISO9001:1994 (the "old" Standard) and ISO/DIS9001:2000 (I-2k) when it comes to the word "trends."

By the length of this post (for which I apologize to those who get all knotted up if it's more than one paragraph), one can probably tell that Joe has hit a nerve...

I was in an outfit implementing ISO9001:1994 and I encountered terminal Neanderthal syndrome, but (interestingly) it was only attached to certain portions of the old Standard.

Illustration: 4.20 Statistical Techniques was buried "closed coffin" before more than 3 people even read it, because this outfit did professional services and "didn't use statistical techniques." (On my Father's grave I swear, those were the exact words I heard, over and over...).

Time and again in the ISO9000:1994 Standard, individual "elements" are specified in such a way that they (in non-Neanderthal terms) fold back upon the entire quality system: (1) Internal Audit must look at the QMS, but then also must examine itself (an audit of Internal Audit); (2) same for Corrective and Preventive Action; and (3) ditto for Training, and so on.

As if terms like billable base, federal overhead rate, attrition and # of active contracts were all brand new terms, this outfit (and thousands like it, I was later to discover) went charging into QMS implementation looking a little bit ( to me ;o) like those bobbing-head dolls you sometimes see in the back window of the car in front of you... operating a services business and implementing a QMS, all the while saying "... we don't use statistical techniques."

So I stand on my feet and APPLAUD the writers of I-2k for several things, the most enjoyable of which (this is my "author" speaking, now) is Clause 8.4 Analysis of data for improvement. Clause I-2k 8.4 is, effectively, old clause 4.20 Statistical techniques, glued to a stone mallet and delivered wide-swing to the forehead of Neanderthals such that they should be extinct, or at least embarrassed.

Because the word "trends" in I-2k is nothing new.

(A.) Under ISO9001:1994 4.14, look at any entity's current list of corrective actions -- try not to group them by ISO-clause or Procedure-clause, because if you did you'd be starting to determine "... the magnitude of problems" (which is detectable as a trend) so that corrective action can be commensurate with the risks.

(B.) Under ISO9001:1994 4.18, "Appropriate records of training shall be maintained." One time I asked the question, "Shouldn't we test after training to see if folks GOT IT?" and the answer I got sounded something like this: "People come to headquarters to eat, and to drink, and not to fail tests." Proof that the word "appropriate" means different things, depending on what rung of the evolutionary ladder you're on. (In fact, another interesting trend I observed was how, each time that it turned out I was correct about this or that disputed interpretation regarding ISO9001:1994, it would suddenly be paraphrased by another senior manager and become a brilliant idea. More bobbing heads.)

(C.) Under ISO9001:1994, there are eighteen "(see 4.16)" quality record requirements. These are all supposed to provide a means for satisfying the 1 SCOPE subparagraph b) requirement that "confidence in product conformance can be attained by adequate demonstration of ... capabilities..." but in order to do that you need to analyze the data and establish trends.

(D.) There are many more "trend" locations in the old Standard, but I think I've made my point !!!

The Editor of McGraw Hill's "Quality Systems Update," J. Paul Scicchitano, summarized the results of a 1999 survey of ISO9k players by stating, "Companies that attempt to exceed the basic requirements of ISO9000 tend to have a more favorable return on investment than those companies that do not."

If your entity elected to exceed the "strict" requirements of ISO9001:1994 and, in fact, did begin to establish and implement statistical data under "old" clause 4.20 and to take a serious look at meaningful trends, then (1) your entity is probably much better off for it; and, (2) your entity has very little adjusting to do for I-2k "trends" because you've already been doing it.

Even to a newbie-RAB auditor, "corporate smoke" will look a lot like bobbing heads and "hey, we're new at this statistics stuff." On the other hand, a mature QMS will quickly rise to the surface based on its many years of trend analysis justification.

But the WAY that I-2k, Section 8 is worded, it sure makes it easy for a Registrar's Auditor (RAB or otherwise) to write a nonconformity report (audit finding, observation, or whatever term you want) where non-effectiveness is detected simply because OBVIOUS data is ignored.

Separates the wheat from the chaff, and quickly amends the bobbing-head's approach to "old" 4.20. Now THAT's continuous improvement!

David M. Kozenko